# Metaphlan database has duplicate entries, causing downstream failures

**URL:** <https://forum.biobakery.org/t/metaphlan-database-has-duplicate-entries-causing-downstream-failures/8188>\
**Category:** MetaPhlAn\
**Created:** [June 26, 2025, 5:16pm UTC](https://forum.biobakery.org/t/metaphlan-database-has-duplicate-entries-causing-downstream-failures/8188 "2025-06-26T17:16:52Z")\
**Posts on this page:** 2\
**Page:** 1

<div class="post-metadata">

**Author:** ![seandavi](https://yyz1.discourse-cdn.com/flex027/user_avatar/forum.biobakery.org/seandavi/32/3304_2.png) [@seandavi](https://forum.biobakery.org/u/seandavi)\
**Post date:** [June 26, 2025, 5:16pm UTC](https://forum.biobakery.org/t/metaphlan-database-has-duplicate-entries-causing-downstream-failures/8188/1 "2025-06-26T17:16:52Z")

</div>

The metaphlan database, `mpa_vOct22_CHOCOPhlAnSGB_202403`, has duplicate sequences, resulting in duplicate `@SQ` headers in SAM output. Bowtie2 errors on those when using a `.sam` file produced by metaphlan4. This duplication seems to be similar to that reported here: [VSG sequence duplications found in MetaPhlAn4 vOct22 pre-built bowtie2 index file](https://forum.biobakery.org/t/vsg-sequence-duplications-found-in-metaphlan4-voct22-pre-built-bowtie2-index-file/5106)

An example duplicated sequence:

```auto
bunzip2 -c mpa_vOct22_CHOCOPhlAnSGB_202403.fna.bz2 | grep 'VDB|003B-0000-0-01C2|M1-c0-c0-c0'
>VDB|003B-0000-0-01C2|M1-c0-c0-c0 240413_kVSG_NC_0274021_Salmonella_phage_SPN3US
>VDB|003B-0000-0-01C2|M1-c0-c0-c0 240413_kVSG_NC_0274021_Salmonella_phage_SPN3US

```

An example error:

```auto
Thu Jun 26 16:35:33 2025: Start samples to markers execution
Thu Jun 26 16:35:33 2025: Using samtools version 1.21
Thu Jun 26 16:35:33 2025: Creating temporary directory...
Thu Jun 26 16:35:33 2025: Done.
Thu Jun 26 16:35:33 2025: Filtering SAM files...
Thu Jun 26 16:35:33 2025: Loading MetaPhlAn mpa_vOct22_CHOCOPhlAnSGB_202403 database...
Thu Jun 26 16:36:00 2025: Done.
Thu Jun 26 16:36:03 2025: Setting default parameters for mapper bowtie2
[E::sam_hrecs_refs_from_targets_array] Duplicate entry "VDB|003B-0000-0-01C2|M1-c0-c0-c0" in target list
[E::sam_parse1] failed to parse header
Traceback (most recent call last):
  File "/usr/local/bin/sample2markers.py", line 8, in <module>
    sys.exit(main())
  File "/usr/local/lib/python3.9/site-packages/metaphlan/utils/sample2markers.py", line 587, in main
    sampletomarkers.run_sample2markers()
  File "/usr/local/lib/python3.9/site-packages/metaphlan/utils/sample2markers.py", line 438, in run_sample2markers
    self.filter_sam_files()
  File "/usr/local/lib/python3.9/site-packages/metaphlan/utils/sample2markers.py", line 423, in filter_sam_files
    self.input = execute_pool(((SampleToMarkers.parallel_filter_sam, i, self.tmp_dir, self.input_format,
  File "/usr/local/lib/python3.9/site-packages/metaphlan/utils/parallelisation.py", line 102, in execute_pool
    return list(gen)
  File "/usr/local/lib/python3.9/site-packages/metaphlan/utils/parallelisation.py", line 95, in <genexpr>
    gen = (function(*a) for function, *a in args)
  File "/usr/local/lib/python3.9/site-packages/metaphlan/utils/sample2markers.py", line 386, in parallel_filter_sam
    aln = pysam.AlignedSegment.fromstring(line, header)
  File "pysam/libcalignedsegment.pyx", line 1124, in pysam.libcalignedsegment.AlignedSegment.fromstring
ValueError: parsing SAM record string failed (error code -2)

```

---

<div class="post-metadata">

**Author:** ![Claudia\_Mengoni](https://yyz1.discourse-cdn.com/flex027/user_avatar/forum.biobakery.org/claudia_mengoni/32/2871_2.png) [@Claudia\_Mengoni](https://forum.biobakery.org/u/Claudia_Mengoni)\
**Post date:** [July 1, 2025, 4:13pm UTC](https://forum.biobakery.org/t/metaphlan-database-has-duplicate-entries-causing-downstream-failures/8188/2 "2025-07-01T16:13:50Z")

</div>

Hi @seandavi  
Thanks for spotting that, the issue was with the viral sequences only, so the rest of the mappings shouldn’t be affected by this. I fixed the files, you could either rerun the analysis with the updated database or remove the VDB hits if your aim is to run sample2markers.  
Thanks!